Various methods have been suggested for performing color reproduction between color image reproducing apparatuses having different color characteristics such as a color scanner, a color monitor, a color projector, a color printer, some of which will be described as an example. Although the color image reproducing apparatus is variously called depending on the literatures cited in the following description, it is called “color image reproducing apparatus” in this specification for the sake of clarification.
In a color reproduction method disclosed in Annex D of Non-Patent literature 1, a correspondence relation between a color device value of a color image reproducing apparatus and the Profile Connection Space (PCS) that is specified by the ICC (International Color Consortium) and has
CIE standard illuminant D50 as reference white is obtained first. Then, a color device value of the color image reproducing apparatus which is an input source is converted to a color device value of the color image reproducing apparatus which is an output destination through the PCS. More specifically, the color device value of the color image reproducing apparatus of the input source is converted into the PCS, and then the PCS is converted into the color device value of the color image reproducing apparatus which is the output destination. When the conversion from the color device value to the PCS is performed, if the whites are different, chromatic adaptation conversion of human complete chromatic adaptation type is applied. The color reproduction in this case corresponds to color conversion in which the color device value corresponding to white of the color image reproducing apparatus which is the input source is converted to the color device value corresponding to white of the color image reproducing apparatus which is the output destination.
A color conversion method, which takes the human complete chromatic adaptation into account, is based on the von Kries model. According to this color conversion method, the chromatic adaptation is carried out based on the change in three kinds of spectral sensitivity (ργβ) in the human visual system such that the spectral sensitivity is upon change in illumination, changed in its sensitivity balance without changes in shape of the spectral curves so as to bring the two whites into coincidence. Assume ργβ sensual quantities of illumination 1 be (ρ0, γ0, β0), the ργβ sensual quantities of an object in illumination 1 be (ρ, γ, β), and ργβ sensual quantities of illumination 2 be (ρ0′, γ0′, β0′), with ργβ sensual quantities of the same object under illumination 2 being (ρ′, γ′, β′) Then, the ργβ sensual quantities of the visual system of the object color are expressed by: ρ/ρ0, γ/γ0, β/β0, ρ′/ρ0′, γ′/γ0′, and β′/β0′. In order for the color appearance of the object to be coincident under the illumination 1 and under the illumination 2, it suffices if the above-mentioned ργβ sensual quantities are coincident, as shown in the formula (1) and the formula (2).
                              [                                                                                          ρ                    /                    ρ                                    ⁢                                                                          ⁢                  0                                                                                                                          γ                    /                    γ                                    ⁢                                                                          ⁢                  0                                                                                                                          β                    /                    β                                    ⁢                                                                          ⁢                  0                                                              ]                =                  [                                                                                                                ρ                      ′                                        /                                          ρ                      ′                                                        ⁢                                                                          ⁢                  0                                                                                                                                                γ                      ′                                        /                                          γ                      ′                                                        ⁢                                                                          ⁢                  0                                                                                                                                                β                      ′                                        /                                          β                      ′                                                        ⁢                                                                          ⁢                  0                                                              ]                                    (        1        )                                                      [                                                                                                      1                      /                      ρ                                        ⁢                                                                                  ⁢                    0                                                                    0                                                  0                                                                              0                                                                                            1                      /                      γ                                        ⁢                                                                                  ⁢                    0                                                                    0                                                                              0                                                  0                                                                                            1                      /                      β                                        ⁢                                                                                  ⁢                    0                                                                        ]                    ⁡                      [                                                            ρ                                                                              γ                                                                              β                                                      ]                          =                              [                                                                                                      1                      /                                              ρ                        ′                                                              ⁢                                                                                  ⁢                    0                                                                    0                                                  0                                                                              0                                                                                            1                      /                                              γ                        ′                                                              ⁢                                                                                  ⁢                    0                                                                    0                                                                              0                                                  0                                                                                            1                      /                                              β                        ′                                                              ⁢                                                                                  ⁢                    0                                                                        ]                    ⁡                      [                                                                                ρ                    ′                                                                                                                    γ                    ′                                                                                                                    β                    ′                                                                        ]                                              (        2        )            
Further, the ργβ sensual quantities can be obtained from the tristimulus values XYZ by linear transformation as shown in the formula (3).
                              [                                                    ρ                                                                    γ                                                                    β                                              ]                =                                            M              ⁡                              [                                                                            X                                                                                                  Y                                                                                                  Z                                                                      ]                                      ⁢                                                  [                                                                                ρ                    ′                                                                                                                    γ                    ′                                                                                                                    β                    ′                                                                        ]                    =                      M            ⁡                          [                                                                                          X                      ′                                                                                                                                  Y                      ′                                                                                                                                  Z                      ′                                                                                  ]                                                          (        3        )            
It is noted that ρ0, γ0, β0, ρ0′, γ0′, β0′ in the above formula (1) may be obtained by substituting tristimulus values (X0, Y0, Z0), (X0′, Y0′, Z0′) of the illumination 1 and the illumination 2 into the formula (3). Further, a matrix of Pitt, a matrix of Estevez, a matrix of Bradford or the like is used as the transformation matrix M in the formula (3).
Further, the von Kries chromatic adaptation prediction formula shown in the formula (4) may be obtained by substituting the formula (3) into the formula (2). Note that D in the formula (4) is shown in the formula (5).
                              [                                                                      X                  ′                                                                                                      Y                  ′                                                                                                      Z                  ′                                                              ]                =                              M                          -              1                                ⁢                      DM            ⁡                          [                                                                    X                                                                                        Y                                                                                        Z                                                              ]                                                          (        4        )                                D        =                  [                                                                                          ρ                    ′                                    ⁢                                      0                    /                    ρ                                    ⁢                                                                          ⁢                  0                                                            0                                            0                                                                    0                                                                                  γ                    ′                                    ⁢                                      0                    /                    γ                                    ⁢                                                                          ⁢                  0                                                            0                                                                    0                                            0                                                                                  β                    ′                                    ⁢                                      0                    /                    β                                    ⁢                                                                          ⁢                  0                                                              ]                                    (        5        )            
In the color image reproducing apparatus of the RGB color system, the correspondence relation between the color device values RGB and the tristimulus values XYZ is as shown in formula (6). It is assumed here that the color device values RGB are obtained by linearly converting the input characteristic by performing inverse gamma correction in advance in consideration of a gamma characteristic of the color image reproducing apparatus.
                              [                                                    X                                                                    Y                                                                    Z                                              ]                =                  Q          ⁡                      [                                                            R                                                                              G                                                                              B                                                      ]                                              (        6        )            
In the formula (6), Q is 3×3 matrix, which will be hereinafter referred to as an RGB-XYZ transformation matrix. The RGB-XYZ transformation matrix Q varies depending on color characteristics of the color image reproducing apparatus. When black tristimulus values (Xk, Yk, Zk) are considered, the formula (6) can be expressed as formula (7).
                              [                                                    X                                                                    Y                                                                    Z                                              ]                =                              Q            ⁡                          [                                                                    R                                                                                        G                                                                                        B                                                              ]                                +                      [                                                                                X                    k                                                                                                                    Y                    k                                                                                                                    Z                    k                                                                        ]                                              (        7        )            
Thus, when the black tristimulus values are zero, the color conversion from the RGB of the color image reproducing apparatus having the RGB-XYZ transformation matrix Q of “A” to the R′G′B′ of the color image reproducing apparatus having the RGB-XYZ transformation matrix Q of “B” is as shown in formula (8), using the above-stated color conversion method considering the human complete chromatic adaptation.
                              [                                                                      R                  ′                                                                                                      G                  ′                                                                                                      B                  ′                                                              ]                =                              B                          -              1                                ⁢                      M                          -              1                                ⁢                      DMA            ⁡                          [                                                                    R                                                                                        G                                                                                        B                                                              ]                                                          (        8        )            
Note that the matrix D in the formula (8) can be calculated using the tristimulus values of white of each of the two color image reproducing apparatuses and the formulae (3) and (5).
Patent literature 1 suggests a color reproduction method considering incomplete chromatic adaptation in color reproduction between color image reproducing apparatuses having different whites. This method restores white surface reflectances of the conversion source and the conversion destination using a correlated color temperature of white of the color image reproducing apparatus of the conversion source and the conversion destination, and calculates, from two white surface reflectances that are restored, white adaptation white surface reflectance in the state of incomplete chromatic adaptation. Then, the relation between the white surface reflectance of the conversion source and the white adaptation white surface reflectance in the state of incomplete chromatic adaptation is similarly applied to the surface reflectance of other colors, thereby achieving coincident color appearance between color image reproducing apparatuses having different whites.
Further, Patent literature 2 discloses an image processing apparatus that executes desired image processing to an input signal using a color correction table for changing an amount of conversion for a white point in accordance with a white luminance provided from a color image reproducing apparatus, to output the signal to the color image reproducing apparatus. This image processing apparatus uses the characteristic that the human eye tends to adapt to bright light more than to dark light. The processing is performed to make a target white point after the conversion close to a white point of the color image reproducing apparatus in accordance with a white luminance provided from the color image reproducing apparatus being increased.
Furthermore, Patent literature 2 discloses an image processing apparatus that is provided with a color correction table which prioritizes color reproduction and is configured to perform color correction for achieving faithful color reproduction, and a color correction table which prioritizes lightness and is configured to perform color correction that prioritizes lightness. Patent literature 2 further discloses a method of making a target white point after conversion by the color correction table that prioritizes color reproduction close to a white point of the color image reproducing apparatus's, and a method of switching the color correction table that prioritizes lightness and the color correction table that prioritizes the color reproduction according to the input signal such as a natural image or presentation data. In switching the tables according to the input signals, the table is switched to the color conversion table that prioritizes the color reproduction when the input signal is natural image data, and the table is switched to the color correction table that prioritizes lightness when the input signal is presentation data.
Further, Patent literature 3 discloses a method of adding, to the method disclosed in Patent literature 2, processing to change an amount of conversion for a white point according to the elapsed time after the start of color correction.
The methods disclosed in Patent literatures 1 to 3 relate to a color reproduction method of setting the chromaticity of the adaptation white reproduced by the color image reproducing apparatus to a desired chromaticity considering the (incomplete) chromatic adaptation in the human color perception and applying the color conversion on the basis of the adaptation white for other colors than white by the corresponding color reproduction or an equivalent color reproduction.
Further, it is relatively easy for the color display to achieve color reproduction by using a specific color space (e.g., CIE standard illuminant D65 in sRGB and Adobe RGB) in which white chromaticity of a color to be reproduced is specified by means of colorimetric color reproduction. Further, the related methods of the above-mentioned Patent literatures 1 to 3 show the color reproduction method of setting the chromaticity of the adaptation white reproduced by the color image reproducing apparatus to a desired chromaticity considering the (incomplete) chromatic adaptation in the human color perception and applying the color correction (color conversion) obtained on the basis of the adaptation white for other colors than white as the corresponding color reproduction or the equivalent color reproduction.